Are there any threads where measuring methods are being discussed?
In the day we'd buy a scope, a function generator, some multimeters, lab supplies and if we felt rich a power meter and distortion analyzer.
But today, with cheap test sources based on PCs and Androids, I'm curious what people are doing at home.
At work we still spend tons of money.... but we now get automated signal "analyzers"... way overkill for home.
I sold my Tektronix scope a long time ago. 😛
Looks like I ain't gettin' nuthin'
Bose Crapbar-Jeesh. Yeah I miss my test equipement. Need a scope but time is limited and have to bend to the times. This is not my field. But the only one this body can deal with now.
"nuthin"- no Iron Pre for you?!
"nuthin"- no Iron Pre for you?!
I think I used these. They don't need to be anything special.
https://www.mouser.com/ProductDetail/Aavid/507302B00000G?qs=asPD7ZL2j3VyLVwHkYVBBw==
Being near the caboose... somethings are good, like we don't get the smoke from the steam locomotive up front... bad things are, we are the last ones to pull into the station.
I was honored to be asked how I think about what gain to choose for a pre-amp, specifically the Iron Pre SE with its variable gain.
Posting pieces of my answer here in case it may help someone else. Most importantly, I'm posting it selfishly to help myself learn and help the other person in case any of it is complete BS / wrong.
The tool I mention is an exceptional piece of wizardry posted by @fireanimal in the Wolverine thread. I've taken his awesome tool and perhaps butchered it for use with other amplifiers.
-----------
The way I have tried to think about it is that I never want either the power amplifier or the pre amplifier to ever voltage clip. It is also nice if the pre amplifier has enough gain and voltage output capability to drive the power amplifier close to clipping, but not over. However, I don't like to have it be able to go over the limit in case someone accidentally turns it all the way to the highest volume. In my mind, a pre-amplifier should NEVER clip with a reasonable music source at its highest output with the source putting out a signal of 0dBFS.
So the maximum voltage output of the pre-amp should always be greater than its gain * 2V (typical rms max output of a source) * 1.414 (convert to Vp).
So the max output of an SE pre with 2x (6dB) gain needs to be greater than 4Vrms or 2 * 2 * 1.414 => 5.7Vp
Max output of an SE pre with 4x (12dB) gain needs to be greater than 8Vrms or 4 * 2 * 1.414 => 11.4Vp
To calculate the required maximum output of a balanced pre, I use 5Vrms as potential input. Most sources will only output 4Vrms, but oddly, I have one that will do 5Vrms.
So the max output of an Balanced pre with 2x (6dB) gain needs to be greater than 10Vrms or 5 * 2 * 1.414 => 14.14Vp
Those were for example, but the Iron Pre will safely do all those input voltages at any of those gains and not voltage clip. It will also do it at more than adequate distortion.
Next... what gain to use. Slightly different angle. I look backwards from the power amplifier. The goal is to be able to 'preserve the signal' and not attenuate as much. Running your volume knob at maximum '9 o'clock' is a waste of gain, but on the opposite end, clipping your amplifier is bad.
---------
Since you have started with the Wolverine (I am building two also), we need to know your voltage rails in order to know perfectly how it will work, but I will show an example.
You would look at tab 3 for some theory you can apply to other amplifiers also. It shows the typical voltage losses of the Wolverine. I think of this as how close can I get my music signal to the rails. How much 'usable' voltage do I have from my power supply to put out to my speakers.
In the case of the Wolverine, it's around 8V of loss. For most FW amplifiers with MOSFET outputs, I typically use ~4V loss.
So, with a +-60V PSU, you have +- 52V => 52Vp => 104Vpp of 'swing'. I still hate that term. I think it confuses more people than it helps. Most importantly, I like to convert to Vrms to make the math easier for certain things also. Vp is important to understand clipping, but the math is easier with Vrms. 104Vpp => 36V7rms
So... you know if either the positive or negative half of the signal hits over 52V, a Wolverine with 60V rails will clip.
The wolverine has 26dB (20x) gain....
52Vp / 20 => 2V6p
or my preferred way... 36.7Vrms / 20 => 1.8Vrms
So, the input sensitivity of a Wolverine with a +- 60V PSU may be 1.8Vrms.
That (to me) is the critical spec we need for this discussion.
Your typical SE source will output 2Vrms.
Do you need pre-amp gain at all?
The answer is maybe, but not too likely.
Try playing with the typical music calculator in the tool and see the effect real music has on the outcome.
It's very early my time, and I just had my first cup of coffee. But, I hope my math was correct and that the theory / math help you to think about this in a new way.
Note ... I still am not sure I think about it correctly... 😉
Posting pieces of my answer here in case it may help someone else. Most importantly, I'm posting it selfishly to help myself learn and help the other person in case any of it is complete BS / wrong.
The tool I mention is an exceptional piece of wizardry posted by @fireanimal in the Wolverine thread. I've taken his awesome tool and perhaps butchered it for use with other amplifiers.
-----------
The way I have tried to think about it is that I never want either the power amplifier or the pre amplifier to ever voltage clip. It is also nice if the pre amplifier has enough gain and voltage output capability to drive the power amplifier close to clipping, but not over. However, I don't like to have it be able to go over the limit in case someone accidentally turns it all the way to the highest volume. In my mind, a pre-amplifier should NEVER clip with a reasonable music source at its highest output with the source putting out a signal of 0dBFS.
So the maximum voltage output of the pre-amp should always be greater than its gain * 2V (typical rms max output of a source) * 1.414 (convert to Vp).
So the max output of an SE pre with 2x (6dB) gain needs to be greater than 4Vrms or 2 * 2 * 1.414 => 5.7Vp
Max output of an SE pre with 4x (12dB) gain needs to be greater than 8Vrms or 4 * 2 * 1.414 => 11.4Vp
To calculate the required maximum output of a balanced pre, I use 5Vrms as potential input. Most sources will only output 4Vrms, but oddly, I have one that will do 5Vrms.
So the max output of an Balanced pre with 2x (6dB) gain needs to be greater than 10Vrms or 5 * 2 * 1.414 => 14.14Vp
Those were for example, but the Iron Pre will safely do all those input voltages at any of those gains and not voltage clip. It will also do it at more than adequate distortion.
Next... what gain to use. Slightly different angle. I look backwards from the power amplifier. The goal is to be able to 'preserve the signal' and not attenuate as much. Running your volume knob at maximum '9 o'clock' is a waste of gain, but on the opposite end, clipping your amplifier is bad.
---------
Since you have started with the Wolverine (I am building two also), we need to know your voltage rails in order to know perfectly how it will work, but I will show an example.
You would look at tab 3 for some theory you can apply to other amplifiers also. It shows the typical voltage losses of the Wolverine. I think of this as how close can I get my music signal to the rails. How much 'usable' voltage do I have from my power supply to put out to my speakers.
In the case of the Wolverine, it's around 8V of loss. For most FW amplifiers with MOSFET outputs, I typically use ~4V loss.
So, with a +-60V PSU, you have +- 52V => 52Vp => 104Vpp of 'swing'. I still hate that term. I think it confuses more people than it helps. Most importantly, I like to convert to Vrms to make the math easier for certain things also. Vp is important to understand clipping, but the math is easier with Vrms. 104Vpp => 36V7rms
So... you know if either the positive or negative half of the signal hits over 52V, a Wolverine with 60V rails will clip.
The wolverine has 26dB (20x) gain....
52Vp / 20 => 2V6p
or my preferred way... 36.7Vrms / 20 => 1.8Vrms
So, the input sensitivity of a Wolverine with a +- 60V PSU may be 1.8Vrms.
That (to me) is the critical spec we need for this discussion.
Your typical SE source will output 2Vrms.
Do you need pre-amp gain at all?
The answer is maybe, but not too likely.
Try playing with the typical music calculator in the tool and see the effect real music has on the outcome.
It's very early my time, and I just had my first cup of coffee. But, I hope my math was correct and that the theory / math help you to think about this in a new way.
Note ... I still am not sure I think about it correctly... 😉
Last edited:
I should have received my letter about an hour ago, but the system is now showing 14 people ahead of me. When I check the event timer, it comes up as a past event (1 hour ago). Is anybody else having the same issue?
Jason may chime in himself, but in the interim...
Try using the tool to resend your e-mail. Post back here.
Edited to add - and check your 'spam folder'
Try using the tool to resend your e-mail. Post back here.
Edited to add - and check your 'spam folder'
Nothing in the spam box.
Not sure how I can get the tool to resend the email. When I check the queue it now says 13 people ahead of me, 10.5 hours to go, but the event timer still says 8pm local time (which was 90min ago).
Not sure how I can get the tool to resend the email. When I check the queue it now says 13 people ahead of me, 10.5 hours to go, but the event timer still says 8pm local time (which was 90min ago).
Thanks very much for posting...
We're giving it a look in real time...
<hums pleasing hold music> 🙂
We're giving it a look in real time...
<hums pleasing hold music> 🙂
Morning Fellas. Apologies if this has been posted; rather long thread. Is there a link to get on the list to order? Also is there pre determined cost? Think ill build this for my pops to round out his complete system change
^ It's too late to get on the list we're working with now. Even if we added you, the chances are... you would not reach the front of the line in time to get one.
Two things though...
1) There is a chance (however limited) that all the initial kits won't sell by the time we reach the end of the list. If they don't ... the remaining quantity will be listed in the store normally.
2) There will be future kits in the store. No, we don't know the timing yet.
Hope your pops loves it!
Edited to add - need more coffee. Cost.... $125 for SE and $150 for balanced assuming costs don't take another massive jump.
Two things though...
1) There is a chance (however limited) that all the initial kits won't sell by the time we reach the end of the list. If they don't ... the remaining quantity will be listed in the store normally.
2) There will be future kits in the store. No, we don't know the timing yet.
Hope your pops loves it!
Edited to add - need more coffee. Cost.... $125 for SE and $150 for balanced assuming costs don't take another massive jump.
Ah gotcha. thanks for the cliff notes version. I'll keep an eye out on the store and if for some reason enough people drop out and you need someone to scoop up one; hit me up. Thanks again
Sorry in a way to have brought this gain problem up. But it has given fruit in how best to deal with it through IAIMH and Mighty Zen.
Android maybe not the way to test. No Docs. There are programs on SourceForge, etc. with more ability to run on micoSCUM V. Now on to decipher this additional info.
Thanks, IAIMH.
More coffee for you!
I can get what's suddenly a problem with gain testing?
any crappy sine sig gene is good enough,semi-decent DMM
Freq 400Hz to 1KHz
put DMM on Vac, connect it to input of black box gain stage, start with volume at 0, set to 100mV (say) at input
move red probe to output node, read what's there, presto
if line stage is in case, no connection to amp
if amp is in case, no speaker connected
any crappy sine sig gene is good enough,semi-decent DMM
Freq 400Hz to 1KHz
put DMM on Vac, connect it to input of black box gain stage, start with volume at 0, set to 100mV (say) at input
move red probe to output node, read what's there, presto
if line stage is in case, no connection to amp
if amp is in case, no speaker connected
Maybe some others are in a similar place...
Many moons ago, I had no idea how anything worked (still don't for the most part) or even the difference between gain and volume. I have my old NAD 2700 on the bench now. It was my brother's, and then it became mine. It's one of the first pieces of gear I ever saw with user controlled gain. When I got it, I just thought they were volume knobs with a fancy professional name "gain". If it weren't for the fact that the 'soft clipping' switch was on by default, we would have destroyed that amp and probably some speakers. It's really not a term that I became familiar with until I started playing here.
Fast forward 30 ish years later to playing in this hobby. Early on, the only 'testing' I did after a build was ... did it play music w/o any REALLY obvious issues. I never checked anything else other than major operating points with a DMM. I trusted that if those were correct, that everything was working as intended. That's true for the VAST majority of cases, but there's all sorts of stuff to diddle with if we want.
I don't think there's a problem with gain testing per se. I just think if anyone else is out there like me... any time you do something new, you think you're risking blowing up your project. Once you do it the first time, it's old hat. I still tremble a bit hooking up scope probes. It takes me 30 mins to make sure I don't tie to the wrong ground or to the wrong place in the circuit etc. People like you and Jim, Nelson and so many others have made this hobby so much more accessible to true dolts like me with no electronics training ... It's a double-edged sword. I get to play with incredible stuff that's dumbed down for me, but I still wonder A LOT when will be the next time I'll release the magic smoke.
That reminds me... I really need to build or buy a set of differential probes.
Many moons ago, I had no idea how anything worked (still don't for the most part) or even the difference between gain and volume. I have my old NAD 2700 on the bench now. It was my brother's, and then it became mine. It's one of the first pieces of gear I ever saw with user controlled gain. When I got it, I just thought they were volume knobs with a fancy professional name "gain". If it weren't for the fact that the 'soft clipping' switch was on by default, we would have destroyed that amp and probably some speakers. It's really not a term that I became familiar with until I started playing here.
Fast forward 30 ish years later to playing in this hobby. Early on, the only 'testing' I did after a build was ... did it play music w/o any REALLY obvious issues. I never checked anything else other than major operating points with a DMM. I trusted that if those were correct, that everything was working as intended. That's true for the VAST majority of cases, but there's all sorts of stuff to diddle with if we want.
I don't think there's a problem with gain testing per se. I just think if anyone else is out there like me... any time you do something new, you think you're risking blowing up your project. Once you do it the first time, it's old hat. I still tremble a bit hooking up scope probes. It takes me 30 mins to make sure I don't tie to the wrong ground or to the wrong place in the circuit etc. People like you and Jim, Nelson and so many others have made this hobby so much more accessible to true dolts like me with no electronics training ... It's a double-edged sword. I get to play with incredible stuff that's dumbed down for me, but I still wonder A LOT when will be the next time I'll release the magic smoke.
That reminds me... I really need to build or buy a set of differential probes.
Last edited:
gain testing we are cringing about on last few pages is nothing else than way of finding/confirming is there any problem with Gizmos made by Holzarbeiter and Iforgotwhoelse
TonyEE asked for TurboGigaMega Iron Pre with 20db of gain (to drive bridged F4); I did show him that existing Iron Pre Bal (with 6db) followed with Edcor PC600:15K (with 14db) is resulting in asked 20db
cheap as few innings in local Starbucks, but tasting much better
TonyEE asked for TurboGigaMega Iron Pre with 20db of gain (to drive bridged F4); I did show him that existing Iron Pre Bal (with 6db) followed with Edcor PC600:15K (with 14db) is resulting in asked 20db
cheap as few innings in local Starbucks, but tasting much better